Table Data Conversion to Pixel Clusters for Financial Automation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current software systems face difficulties in recognizing and processing financial data presented in tabular formats, requiring manual review and input by accounting advisors, which is time-consuming and costly.

Innovation Solution

A data processing system that digitally recognizes tables by converting source data into machine-encoded text data, pixelating it, clustering similar data, and classifying it into rows and columns, using techniques like density clustering and optical character recognition to automate the extraction and processing of financial information.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If manual review and input methodology is used, then data processing accuracy is maintained, but processing time and cost increase significantly

Engineering Contradiction:

Improveprocessing speedVSAvoidautomation level

Core Design Contradiction:

ProductivityVSExtent of automation

Solution Approach 1:

The patent replaces manual mechanical processing (accounting advisors reviewing and inputting data) with an automated optical character recognition system that uses image processing and pattern recognition algorithms to extract and process tabular financial data automatically

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service processing where the software automatically recognizes, extracts, and processes tabular data without requiring human intervention, allowing the system to serve itself in completing the data processing task

Inventive Principle:

Principle #25Self-service

2Productivity

If digital processing of table data is implemented, then processing efficiency is improved, but recognition accuracy deteriorates due to difficulty in recognizing table structure

Engineering Contradiction:

Improveprocessing efficiencyVSAvoidtable structure recognition accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent segments the table recognition process into distinct stages: detecting table boundaries, identifying row and column structures, extracting individual cells, and processing content separately, thereby improving overall recognition accuracy through systematic breakdown of the complex task

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces intermediary processing steps including image preprocessing, edge detection, and structural analysis as intermediate stages between receiving the table image and extracting the final data, serving as mediators that enhance recognition accuracy

Inventive Principle:

Principle #24Intermediary (Mediator)

3Ease of operation

If conventional software systems are used, then system simplicity is maintained, but ability to process tabular data deteriorates

Engineering Contradiction:

Improvesystem simplicityVSAvoidtable data processing capability

Core Design Contradiction:

Ease of operationVSAdaptability or versatility

Solution Approach 1:

The patent creates a universal processing system that can handle multiple types of tabular formats and financial documents through a single integrated platform, enabling the system to perform both simple recognition and complex table structure analysis with one unified tool

Inventive Principle:

Principle #6Universality (Multi-functionality)

Data Source

PatentUS11216659B2Converting table data into component parts

Publication Date: 2022.01.04 KPMG LLP

AI summary

A system and method of processing source data that includes table data by converting the table data into machine encoded text data having associated therewith text coordinate data having a Y-axis component and an X-axis component, and then generating from the machine encoded text data a plurality of pixels along the Y-axis component and the X-axis component. The system then performs a clustering technique on the plurality of pixels to generate a plurality of clusters of pixels based on similar attributes, and classifying each of the plurality of clusters of pixels as a selected row of the table and as a selected column of the table, thus making available the information encoded in the table for subsequent processing.